03 · What You Need to Know
What Belongs in a Qualitative Analysis Plan
Identify the Analytical Approach and Why It Fits the Question
Begin by naming and justifying the analytical approach. This matters because qualitative analysis is not one generic procedure.
Thematic analysis, grounded theory, qualitative content analysis, interpretative phenomenological analysis, narrative analysis, discourse analysis, framework approaches, and other methods have different purposes and assumptions. Even within a named family of methods, researchers may make substantially different choices about theory, coding, researcher subjectivity, and what constitutes an analytical finding.
The plan should therefore explain why the selected approach fits the research question, methodological orientation, type of data, and intended form of interpretation. Naming qualitative software is not a substitute for making that argument.
Define What Material Will Enter the Analysis
Specify what counts as analyzable material. Interviews may produce audio recordings, transcripts, interviewer notes, participant-generated materials, and researcher reflections. Observational studies may generate field notes, photographs, documents, recordings, or artifacts. A project may also draw on several sources.
The plan should distinguish which materials constitute primary data and how supplementary materials contribute to interpretation.
This becomes especially important when the final analysis makes claims based on contextual observations or researcher notes that were not described as part of the analytical corpus at the beginning of the study.
Specify How Raw Material Will Be Prepared
Analysis depends partly on what survives the transition from an encounter or source into analyzable material.
For recorded interviews or focus groups, decide how transcription will be handled. Will transcription be verbatim? Will pauses, nonverbal expressions, overlapping speech, or other features matter for the selected analytical approach? How will transcripts be checked against recordings? Will transcripts be translated, and if so, how will translation decisions be managed?
For observational material, determine how field notes will be produced, expanded, dated, and linked to particular settings or events. For documents or online materials, establish inclusion and organization procedures.
These are not merely clerical choices. Different forms of preparation preserve different kinds of information.
Plan Data Organization, Security, and Version Control
Qualitative analysis can generate numerous files and successive versions of transcripts, coding frameworks, memos, diagrams, and analytical outputs. A workable plan should establish how those materials will be organized.
Depending on the project and applicable requirements, this may include participant identifiers, file naming, de-identification, secure storage, access controls, backup procedures, version control, and links among recordings, transcripts, field notes, and analytical documents.
Ethics approvals, institutional requirements, consent agreements, and data-protection rules may impose additional obligations. These should be verified for the particular study rather than assumed to be universal.
Decide When Analysis Will Begin
Do not automatically assume that analysis begins after the last participant leaves.
Many qualitative methodologies involve overlapping data collection and analysis. Early engagement with the material can identify issues that require deeper questioning, cases that challenge emerging interpretations, or areas where additional sampling would be useful.
If analysis will occur concurrently with collection, specify how that interaction will work. If the design intentionally separates the phases, explain that as well.
This is one reason qualitative analysis benefits from planning before data collection. The timing of analysis can affect how the study itself develops.
State Whether the Analytical Orientation Is Inductive, Deductive, or Combined
The plan should indicate where analytical concepts are expected to come from.
In a primarily inductive approach, codes, categories, or patterns may be developed through engagement with the data. In a deductive approach, theory, prior research, a conceptual framework, policy categories, or predefined questions may provide an initial analytical structure. Many studies combine these orientations.
Inductive orientation
Analytical concepts are developed primarily through engagement with the collected material.
Deductive orientation
Analysis begins substantially from concepts, categories, questions, or theoretical ideas established before detailed engagement with the data.
Neither orientation is inherently superior. The choice should reflect the research question and methodology, and the final report should accurately describe what researchers actually did.
Explain How Initial Coding or Interpretation Will Begin
If coding is part of the selected approach, specify how researchers will begin generating and applying codes.
Will researchers code line by line, by meaningful segment, by event, or using another unit? Will an initial codebook be available? If so, where do its codes come from? Will researchers begin by coding a subset of transcripts and then refine the framework?
Not every qualitative methodology conceptualizes analysis primarily as coding. In those cases, the plan should describe the appropriate interpretive procedure rather than forcing the study into a generic code-to-theme sequence.
Specify How Codes, Categories, or Interpretations Can Develop
A plan does not need to contain the final codebook before the data exist. It should, however, explain how the analytical framework will develop.
Codes may be added, renamed, merged, divided, redefined, or removed as researchers compare material. Categories may become more abstract. Relationships among concepts may emerge. Early interpretations may be reconsidered when later material contradicts them.
Planning this process makes change traceable without treating change as methodological failure.
Decide How Comparisons Will Be Made
Comparison is central to many forms of qualitative analysis, although what is compared varies.
Researchers may compare participants, cases, settings, time periods, groups, incidents, codes, categories, or contradictory accounts. Some methodologies use constant comparison explicitly; others use matrices, case summaries, narrative structures, or thematic comparisons.
If the research question involves differences across groups or contexts, the sampling and analysis plans should make those comparisons possible. Do not discover at the end that the study asks about contrasting perspectives but collected almost no material from one of the groups.
Plan How Contradictory and Unexpected Material Will Be Handled
Qualitative analysis should not become a search only for excerpts that support an emerging interpretation.
The plan can establish an expectation that researchers will actively examine material that complicates, contradicts, or falls outside developing patterns. Depending on the methodology, this may involve negative-case analysis, deviant cases, disconfirming evidence, or a broader practice of interrogating variation.
The objective is not to force every account into one coherent story. Sometimes disagreement and variation are themselves important findings.
Plan the Use of Memos and Other Analytical Records
Analytic memos provide a place to record developing interpretations, questions, relationships, uncertainties, comparisons, and decisions. They can help distinguish the evolution of analysis from the final polished account.
The plan might specify when memos will be written, where they will be stored, how they will relate to codes or cases, and whether they will be discussed within the research team.
Other useful records may include codebook versions, diagrams, matrices, meeting notes, case summaries, or decision logs. The appropriate documentation depends on the analytical approach.
Build Reflexivity Into the Analytical Process
Qualitative researchers participate in producing and interpreting the material they study. Their assumptions, disciplinary backgrounds, relationships with participants, theoretical commitments, identities, and positions may shape what they ask, notice, and interpret.
COREQ, for example, includes reporting items concerning researcher characteristics, credentials, occupation, experience, training, relationships with participants, and researcher characteristics that may influence the research. These reporting expectations reinforce the importance of considering researcher position during the study rather than reconstructing it only when writing the manuscript.
The analysis plan may therefore include reflexive journals, memos, team discussions, positionality work, or other practices consistent with the methodology.
Define Roles When Analysis Is Conducted by a Team
When several researchers participate, specify who will do what.
Will all researchers code the same material initially? Will analysts divide cases and then compare interpretations? Who maintains the coding framework? How will meetings be conducted? How will disagreements or alternative interpretations be documented and resolved, if resolution is appropriate?
Do not assume that numerical intercoder agreement is required in every qualitative methodology. Some approaches treat coding consistency as important; others understand coding as interpretive and use multiple researchers to enrich reflexive discussion rather than to establish one objectively correct coding decision.
Specify How You Will Move From Codes to Higher-Level Interpretation
A frequent weakness in qualitative methods descriptions is considerable detail about coding followed by almost no explanation of how researchers reached themes, categories, theoretical propositions, narratives, or other interpretive claims.
The plan should address that transition.
Depending on the approach, researchers may group related codes, compare cases, develop conceptual categories, construct matrices, examine relationships, write thematic summaries, develop theoretical propositions, or repeatedly move between parts of the data and the developing interpretation.
Coding is often an analytical tool. It is not necessarily the final analytical product.
Decide How Interpretations Will Remain Grounded in the Material
The plan should explain how the research team will examine the relationship between analytical claims and the underlying data.
This may involve revisiting full transcripts rather than relying only on coded fragments, examining cases that do not fit, comparing interpretations across analysts, maintaining links from themes to source material, or using other procedures appropriate to the methodology.
Reporting guidelines such as COREQ encourage researchers to provide supporting quotations and indicate the consistency between presented data and findings. The exact evidentiary practice differs among qualitative traditions, but the broader principle is that interpretive claims should remain traceable to the material from which they were developed.
Decide What Will Remain Flexible
A qualitative analysis plan should explicitly recognize decisions that cannot or should not be finalized in advance.
Final codes, categories, themes, theoretical relationships, decisions about pursuing particular emerging ideas, and aspects of iterative sampling may appropriately evolve in some methodologies. Other decisions, such as the overall analytical tradition, data-security procedures, and basic relationship between the research question and analysis, may be much less flexible.
Making that distinction in advance helps researchers preserve legitimate analytical flexibility without allowing “emergence” to become a blanket justification for undocumented methodological changes.
